图割算法在相位解缠中的应用

Application of Maxflow/Mincut Algorithm in Phase Unwrapping

相位解缠一直以来是干涉测量领域中的一个重要研究方向。传统的相位解缠算法的解缠结果易受到噪声或者截断相位的影响。为了解决上述问题,提高解缠精度,在模拟的存在截断相位缺陷的数据上,建立马尔科夫能量模型,推导出能量函数,使得相位解缠变成一个求解全局最优化的问题,利用图割理论求解。实验结果表明,图割理论能够很好的完成能量函数的优化,解缠结果在抗噪性以及精度上,比起传统的解缠算法都有着一定优势。那么就意味着,该方法在相位解缠方面有着重要的研究价值和宽阔的应用前景。

Phase unwrapping is an important field in interference measurement. The traditional phase unwrapping algorithms are easily affected by noise or discontinuous phase. In order to solve the problems and improve solution accuracy, establishing markov energy model, getting the energy function, making the phase unwrapping into a global optimization problem on the datas of simulation of discontinuous phase, using the graph cuts solve the problem. The experimental results show that the optimization of energy function has a good global solution by the theroy, compared with the traditional algorithms, the results has an advantages in noise resistance and precision. That is to say, this method has important research value and application prospect in phase unwrapping.